Cigarette smoking has been called the most important preventable cause of death and is estimated to lead to over 40,000 deaths each year. Recent data illustrate an alarming upward trend in smoking rates among those in their high school years. This is particularly important because data show that of adults who had ever smoked daily, 89 percent has tried a cigarette before the age of 18 and 71.2 had begun smoking daily by age of 18. Moreover, the onset of smoking at a young age is associated with a higher subsequent risk of heavy smoking and of smoking-attributable mortality. The principal aim of the proposed research is to develop policy relevant inforamtion about the responsiveness of youth smoking to changes in tobacco excise taxes. To accomplish this, the project will conduct econometric analyses of the determinants of youth cigareet demand and the onset of smoking. The project wil utilize an exceptionally rich longiduinal secondary data set, the National Educational Longitudianl Study of 1988 (NELS:88). NELS:88 is a large-scale, nationally Representative study providing a variety of data on American 8th graders in 1988, with follow-ups of the same students in 1990, 1992, and 1994. Four distinct advantages of the NELS:88 data allow for innovative research in this area: its longitudinal nature; the inclusion of detailed cigareete consumption measure; availability of a rich set of socioeconomic, school, and parent characteristics; and the availability of all of this information on both high school graduates and dropouts. The econometric investigation consists of two models. Model I consists of cigarette demand functions for 8th, 10th, and 12th graders. In this model the three years of data from NELS:88 will essentially be treated as three separate cross-sections. The empirical results willprovide important information on the responsiveness of youth smoking to cigarette excise taxes and how this responsiveness changes during their high schoool years. The project will conduct separate econometric analyses on the sub-samples of youths who remain in school versus those youths who eventually drop out. Model 2 is an econometric investigation of the determinants of the onset of smoking. This part of the study exploits the fact that NELS:88 follows the same students over time, providing longitudinal data. Alternative measures will be created of the onset of any smoking, daily smoking, and heavy smoking. In addition, the analysis will analyze onset between 8th and 10th grade, and onset between 8th and 12 grade. The central focus on the analysis will be on whether high and increasing cigarette excise taxes decrease the probability of smoking onset for high school students and eventual dropouts. In both Model I and Model 2 the multivariate analysis will include measures of state laws restricting yough access to cigarettes, as well as a wide variety of characteristics of the student respondent, the respondent's parents, teachers, and school characteristics that was collected in each year of NELS:88.